Certificate of Analysis — Guide and Interpretation¶
Document Code: HJ-QA-030
Version: 2.1
Effective Date: 2026-03-10
Prepared by: Finished Product Laboratory & Customer Technical Support, Hongji Agriculture Co., Ltd.
1. Introduction¶
The Certificate of Analysis (COA) is the primary document through which Hongji Agriculture communicates the quality status of each production lot to its B2B customers. A properly designed and accurately completed COA enables the customer's receiving quality team to make rapid, informed decisions about lot acceptance, storage, and process integration.
This guide provides: - A detailed explanation of the COA standard template and each parameter - The analytical method referenced for every test - The specification range and how measured values should be interpreted - Guidance for buyers on using COA data in procurement decisions - Hongji's batch consistency evaluation methodology - Overview of the electronic COA (eCOA) system
2. COA Standard Template¶
Below is the complete COA template used for all Hongji potato flake and potato powder shipments. Each field is described in detail in subsequent sections.
2.1 Template Structure¶
─────────────────────────────────────────────────────────────────────
HONGJI AGRICULTURE CO., LTD.
CERTIFICATE OF ANALYSIS
COA No.: HJ-COA-2026-XXXXX Issue Date: YYYY-MM-DD
Product Name: Potato Flakes (Standard) Grade: Premium
Customer: [Customer Name]
PO No.: PO-XXXXXX-XXXX Lot No.: HJ-AT-R2-260715-A
Production Date: 2026-07-15 Expiry Date: 2027-07-14
Net Weight: 20 kg × 500 bags = 10,000 kg
Pallet IDs: P260715-01 through P260715-20
─────────────────────────────────────────────────────────────────────
PARAMETER UNIT SPECIFICATION RESULT STATUS
─────────────────────────────────────────────────────────────────────
PHYSICAL PROPERTIES
Moisture % 6.0 – 8.0 6.8 PASS
Particle Size (D50) μm 700 – 1200 920 PASS
Particle Size (D90) μm ≤ 2000 1530 PASS
Loose Bulk Density g/mL 0.40 – 0.55 0.47 PASS
Tapped Bulk Density g/mL 0.50 – 0.65 0.58 PASS
Color L* — ≥ 87.0 89.3 PASS
Color a* — −2.5 to +1.5 −1.2 PASS
Color b* — ≤ 22.0 17.5 PASS
Whiteness (WG) — ≥ 78.0 81.2 PASS
Dark Particles per 250 g ≤ 15 3 PASS
CHEMICAL PROPERTIES
Dry Matter % 92.0 – 94.0 93.2 PASS
Starch (% DM) % 70.0 – 80.0 74.8 PASS
Reducing Sugars % ≤ 3.5 1.2 PASS
Protein (N × 6.25) % DM 5.5 – 8.5 6.4 PASS
Ash % DM ≤ 4.5 3.2 PASS
SO₂ Residue mg/kg ≤ 400 86 PASS
MICROBIOLOGICAL PROPERTIES
APC CFU/g ≤ 5.0 × 10⁴ 1.2 × 10³ PASS
Coliforms CFU/g ≤ 5.0 × 10² <10 PASS
E. coli CFU/g ≤ 1.0 × 10² <10 PASS
Salmonella /25g Absent Absent PASS
S. aureus CFU/g ≤ 1.0 × 10³ <10 PASS
Yeasts CFU/g ≤ 5.0 × 10² <10 PASS
Molds CFU/g ≤ 5.0 × 10² 25 PASS
FUNCTIONAL PROPERTIES
Rehydration Ratio g/g ≥ 7.0 7.8 PASS
─────────────────────────────────────────────────────────────────────
DISPOSITION: ✅ APPROVED FOR RELEASE
Verified by: [Name], QA Supervisor Date: YYYY-MM-DD
Digitally Signed: [SHA-256 Hash]
─────────────────────────────────────────────────────────────────────
3. Parameter-by-Parameter Interpretation¶
3.1 Physical Properties¶
| Parameter | Analytical Method | What It Tells the Buyer | Red Flags |
|---|---|---|---|
| Moisture (%) | AOAC 934.06 (oven drying) | Determines yield, shelf life, microbiological stability. High moisture (>8.5%) = shorter shelf life, potential mold; low moisture (<5.5%) = overdried, may have reduced RR | >8.5%: request extended shelf-life data; <5.5%: test rehydration |
| Particle Size D50 (μm) | Sieve analysis (HJ-QA-010) | Median particle diameter indicates texture and hydration behavior. Higher D50 = coarser texture; lower = smoother | Values outside spec may cause clumping, uneven mixing, or off-spec mouthfeel in customer's product |
| Particle Size D90 (μm) | Sieve analysis | Maximum particle size. High D90 indicates oversized particles that may cause gritty texture | >2500 μm: may require re-milling by buyer |
| Bulk Density (Loose/Tapped) | ASTM D1895 | Influences packaging density, transport cost, and volumetric dosing. Low loose BD = high air entrapment = more packaging volume per kg | Significant deviation from historical average may indicate altered process conditions |
| Color (L*a*b*) | CIELAB (D65/10°) | Visual consistency indicator. L* reflects whiteness; a* indicates browning; b* measures yellowness. Critical for brand-consistent finished product | ΔE > 3.0 from buyer's reference: may cause visible difference in finished product |
| Dark Particles | Manual sorting (HJ-QA-010) | Visible quality metric. Higher counts = raw material quality issues or process debris | >20/250g: often triggers customer complaint |
| Whiteness (W_G) | Ganz whiteness formula | Important for white-sauce, beverage, and bakery applications | <75: may produce off-white final product color |
3.2 Chemical Properties¶
| Parameter | Analytical Method | What It Tells the Buyer | Red Flags |
|---|---|---|---|
| Dry Matter (%) | AOAC 934.06 | Inverse of moisture; 92–94% DM means 6–8% moisture | <91.5% DM: microbial stability risk |
| Starch (% DM) | ISO 10520 (Ewers) | Core functional component. Determines thickening power, water binding. | <68% DM: reduced yield in customer's thickening application |
| Reducing Sugars (%) | DNS colorimetric | Primary driver of Maillard browning. High reducing sugars = darker fries/snacks, shorter fry oil life | >3.5%: increased browning risk for customers producing fried products |
| Protein (% DM) | Kjeldahl (N × 6.25) | Minor but relevant for nutritional labeling and Maillard reactivity | Values outside expected range may require nutritional label adjustment |
| Ash (% DM) | AOAC 923.03 (muffle furnace) | Mineral content indicator. High ash may indicate soil contamination or high mineral water | >5.0% DM: investigate soil/residue issue |
| SO₂ Residue (mg/kg) | Modified Monier-Williams | Preservative for color retention. Must meet regulatory limits of destination country | >400 mg/kg: may violate EU/China regulations; >10 mg/kg requires label declaration in US |
3.3 Microbiological Properties¶
| Parameter | Analytical Method | What It Tells the Buyer | Red Flags |
|---|---|---|---|
| APC (CFU/g) | ISO 4833-1 | General hygiene indicator. Low count = good process sanitation | >1 × 10⁵ CFU/g: marginal shelf life; investigate process hygiene |
| Salmonella (/25g) | ISO 6579-1 | Critical pathogen. Zero tolerance in most markets | Detection triggers immediate recall evaluation |
| Yeasts & Molds | ISO 21527-1 | Storage stability indicator. >500 CFU/g may signal moisture ingress or packaging failure | >1 × 10³ CFU/g: increased spoilage risk during transport/storage |
3.4 Functional Properties¶
| Parameter | Analytical Method | What It Tells the Buyer | Red Flags |
|---|---|---|---|
| Rehydration Ratio (g/g) | HJ-QA-025 (70°C, 5 min) | Core functional parameter. Determines yield in mashed potato reconstitution and water binding in formulations | <7.0: poor water absorption; may require recipe reformulation by buyer |
4. How COA Data Guides Buyer Procurement Decisions¶
4.1 Lot Acceptance Decision Matrix¶
Buyers should evaluate COAs against a three-tier decision framework:
| Criterion | Green — Accept | Yellow — Conditional | Red — Reject |
|---|---|---|---|
| All parameters within spec | ✓ | — | — |
| 1–2 minor parameters marginally outside spec (e.g., color b* = 23.5 vs. ≤22.0) | — | Accept with deviation agreement or discounted price | — |
| Any pathogen positive (Salmonella, confirmation required) | — | — | Reject immediately |
| Moisture > 9.0% | — | — | Reject |
| SO₂ > regulatory limit for destination | — | — | Reject |
| RR < 6.0 | — | — | Reject |
4.2 Trend Analysis for Strategic Buying¶
The most sophisticated buyers track COA data over time to build supplier performance profiles:
| Metric | Calculation | Signal |
|---|---|---|
| Cpk (Process Capability Index) | (USL − μ) / 3σ or (μ − LSL) / 3σ (whichever is smaller) | Cpk ≥ 1.33 = capable process; 1.0–1.33 = marginally capable; <1.0 = process improvement needed |
| Ppk (Process Performance Index) | Same formula using overall standard deviation | >1.67 = excellent supplier; customer audits likely low priority |
| Standard Deviation (σ) of key parameters | Across the last 20+ lots | Lower σ = more consistent supplier; enables tighter customer spec |
| Z-Score for each parameter | (Individual value − Batch mean) / σ |
Example — Cpk analysis for 25 consecutive lots of Standard Flakes:
| Parameter | Target | USL | LSL | Mean | σ | Cpk | Rating |
|---|---|---|---|---|---|---|---|
| Moisture (%) | 7.0 | 8.0 | 6.0 | 6.9 | 0.35 | 1.05 | Marginal |
| RR (g/g) | 7.5 | — | 7.0 | 7.7 | 0.15 | 1.56 | Excellent |
5. Batch Consistency Evaluation¶
5.1 Within-Lot Variability¶
For each production lot, Hongji tests 3 composite samples (beginning, middle, and end of the run). The within-lot coefficient of variation (CV) is reported:
$$CV = \frac{\sigma_{within-lot}}{\bar{x}} \times 100\%$$
| Parameter | Typical CV (%) | Acceptable CV (%) | Alert CV (%) |
|---|---|---|---|
| Moisture | 1.0–2.0 | ≤3.0 | >3.0 |
| Reducing Sugars | 2.0–5.0 | ≤8.0 | >8.0 |
| Rehydration Ratio | 1.5–3.0 | ≤5.0 | >5.0 |
| Bulk Density | 1.0–2.5 | ≤4.0 | >4.0 |
5.2 Between-Lot Variability (Rolling 20-Lot Window)¶
Hongji's internal target for between-lot CV is ≤5% for all critical parameters. A rolling window of the last 20 lots is continuously monitored.
Monthly reports showing between-lot trends are available upon request for qualified customers.
6. Electronic Certificate of Analysis (eCOA) System¶
6.1 System Overview¶
Hongji has implemented a fully digital eCOA system, allowing customers to access, download, and integrate quality data into their own systems.
| Feature | Capability |
|---|---|
| Portal Access | Secure customer portal: https://qa.hjpotatoflakes.com |
| API Integration | RESTful API for automated COA retrieval (JSON / XML) |
| Real-Time Availability | COA published within 2 hours of lot release |
| Historical Archive | All COAs retained for minimum 5 years |
| Digital Signature | SHA-256 signed PDF with visible QR code for verification |
| Multi-Format Export | PDF, CSV, XML, JSON |
6.2 QR Code Verification¶
Every Hongji eCOA includes a QR code containing: - Lot number - COA number - SHA-256 hash of the COA content - Public verification URL
Buyers can scan the QR code or visit the verification URL to confirm the COA authenticity.
6.3 API Specification Summary¶
| Endpoint | Method | Description |
|---|---|---|
/api/v1/coa/{lot_number} |
GET | Retrieve COA by lot number |
/api/v1/coa/list?from=YYYY-MM-DD&to=YYYY-MM-DD |
GET | List COAs by date range |
/api/v1/coa/verify |
POST | Verify COA authenticity (submit SHA-256 hash) |
API access requires pre-registration and issuance of API credentials by Hongji's IT department.
7. Common Customer Questions About COA Data¶
| Question | Answer |
|---|---|
| "My lab's result differs from the COA — which do I trust?" | Inter-laboratory variation is normal (up to ±5% for chemical parameters, ±0.5 log for microbiological). Hongji participates in FAPAS proficiency testing (avg. z-score = 1.3). Discrepancies > ±10% should trigger a joint investigation. |
| "Can you add tests to the COA?" | Yes. Custom COA formats can be arranged during contract negotiation. Additional tests may incur a fee. |
| "Do you provide a COA for each pallet?" | Standard: one COA per production lot (up to 20 MT). For contracts requiring per-pallet COAs (e.g., pharmaceutical-grade food ingredients), add a surcharge. |
| "How long after production is the COA available?" | Within 24 hours (standard) or within 2 hours (eCOA portal). |
8. Quality Commitment¶
Hongji Agriculture stands behind every COA we issue. If a lot fails to meet the documented specifications and the deviation is confirmed by a mutually agreed third-party laboratory, we will:
- Accept return of non-conforming goods at our cost
- Issue a credit or replacement lot within 10 working days
- Conduct a full root cause investigation and share findings
9. References¶
- ISO 9001:2015 — Clause 8.2.3: Review of requirements for products and services
- ISO 22000:2018 — Clause 8.7: Control of nonconforming outputs
- FSSC 22000 Scheme Document — Part 3: Certification Requirements
- GB/T 15000.3: Statistical methods for use in proficiency testing
- Hongji Internal Document HJ-QA-COA-SOP-01: COA Generation and Management Procedure
End of Document
References¶
- AOAC International. Official Methods of Analysis, 21st edition. Gaithersburg, MD.
- ISO (International Organization for Standardization). Horizontal methods for food microbiology and physicochemical analysis.
- ASTM International. Standard test methods for particle size and bulk density.
- FSSC 22000. Food Safety System Certification standard.
- Hongji Agriculture Technology Co., Ltd. 2026. "Certificate of Analysis — Guide and Interpretation." Hongji Agriculture Knowledge Center.
This document is part of the Hongji Agriculture (弘基农业) Technical Documentation Series. For more information about our vertically integrated potato supply chain — from seed breeding and cultivation to processing and global export — visit our official B2B website: https://hjpotatoflakes.com